Interactions between plants and insects can span the range between antagonistic, neutral, and mutualistic.  Abiotic conditions have the potential to alter these plant-insect interactions, both in strength and relative contribution.  The annual herb Impatiens capensis (Balsaminaceae) interacts with a suite of insects, including pollinators, a group of generalist herbivores, and a specialist galling insect that attacks flower buds.  Sunlight and nutrients, two environmental variables that are important for I. capensis, were manipulated during summer 2008 using fertilizer (NPK) addition and canopies that were either clear or painted to alter the red:far red light ratio to mimic foliage shade. 

Results/Conclusions

Visits by pollinators were almost twice as common to fertilized compared to unfertilized plants, and individual pollinators visited more flowers once they arrived at fertilized plants.  However, very little of the increased pollinator visitation was due to the most common pollinators – Bombus spp., Apis mellifera, and a small bee species.  Of all of the potential pollinators observed visiting I. capensis, Bombus spp. is considered to be the most effective pollinator; most of the other species are too small to appropriately contact the reproductive structures.  Thus, the increased visitation to fertilized plants was primarily from floral visitors that may be nectar thieves rather than effective pollinators.  Herbivory, both generalist and specialist, also increased on fertilized plants.  The amount of leaf area consumed by generalists increased by more than 25% on fertilized plants.  Similarly, fertilized plants had both more flower bud galls and an earlier date of gall appearance.  Together, these results suggest that there are several detrimental effects of high soil nutrient content.  Despite these results, however, fertilized plants still grew larger and produced more fruits with equivalent seed weight and seed number per fruit.  Thus, I. capensis reproduction may be more limited by nutrients than by antagonists or cheating mutualists.  Light quality had very little effect on plant growth and reproduction or plant interactions with insects, which is surprising considering how important this factor has been for I. capensis growth in other studies and for plant-insect interactions in many systems.  Overall, this study demonstrates that abiotic factors can alter plant-insect interaction strength and the relative contribution of mutualist compared to neutral or antagonistic interactions.  These results support evidence from other systems that abiotic factors can structure communities and alter the geographic mosaic of selection due to changes in species interactions.